Two rounds – both technical.

It took one month to schedule the second round.

Questions were asked on Machine Learning, Deep Learning, NLP, and Generative AI.

1. What is a support vector in SVM?
2. Explain KNN and how do you model it?
3. Do you train KNN?
4. What is a 300B parameter in an LLM?
5. What are the types of libraries you have worked on in NLP?
6. Explain RAG.
7. What is the difference between deep learning and machine learning?
8. What is CNN?
9. What are stemming and lemmatization, and which would you prefer?
10. Let's say that a person chooses stemming over lemmatization, what might be the reason?
11. Why do you use an LLM model if you can fetch the required documents from a vector database in a RAG approach?